Processing and mechanical properties of pure mg and in situ aln reinforced mg 5al composite 3

In-situ formation of Aluminium Nitride composite powder


46
Chapter 3
In-situ formation of Aluminium Nitride composite powder

3.1 Introduction
AlN has been known as a very important ceramic material in many applications due to
its high thermal conductivity, high electrical resistivity, lower thermal expansion
coefficient than alumina ceramics, low dielectric constant, good thermal shock
resistance and good corrosion resistance [1-6]. AlN can be considered to be an ideal
substrate or packaging material for semiconductors [7] and also an attractive material
for making LEDs in the ultraviolet region, electron-emitting devices and high-
temperature electronic devices due to its wide energy band gap (~6.2 eV) [8].

Commercial AlN powder is usually produced by carbothermal reduction of Al
2
O
3
or
Al(OH)
3
and direct nitridation of aluminium [1,9-11]. In the carbothermal method, the
starting material is reduced by carbon and reacted with N
2
at high temperatures
(1300°C). In the direct nitridation method, Al metal powder reacts with N
2
or NH
3

at
high temperatures (1150°C). It is often prepared by methods such as DC arc-
discharge [12], sublimation recrystallization [13,14], sodium flux [15], volatilization-
condensation of AlN powder [16], chemical vapor deposition [17] and combustion
synthesis or self-propagating high-temperature synthesis [18, 19].

MM technique has been mainly used for preparation of materials which are difficult to
be prepared by the ordinary melting methods, for example, oxide dispersion
strengthened alloys and amorphous or supersaturated solid solution alloys with
In-situ formation of Aluminium Nitride composite powder


47
remarkably extended solid solubilities. This method has been extended to use for low
temperature synthesis of AlN by direct nitridation of Al metal in nitrogen and dry
ammonia atmosphere under pressure of 600 kPa through mechanochemical reaction
[20,21]. Since the formation of AlN from Al metal occurs by an exothermic reaction, it
proceeds spontaneously at higher temperatures due to the heat generated by the
nitridation [22].



(a) (b)
Figure 3.1 (a) Morphology of Al powder and (b) structure of pyrazine.

In this study, in-situ formed AlN composite powder is obtained by milling the mixture
of elemental Al with pyrazine (C
4
H
4

N
2
) which shows a ring-type organic material as
shown in Figs. 3.1(a) and (b). Due to the high concentration of nitrogen in parazine,
relatively fast direct nitriding reaction compared to NH
3
gas can be maintained during
the milling process. During the ball milling process, the mechanical energies
transferred to the powder mixture in the presence of highly reactive transition metal
surfaces accelerates the decomposition of pyrazine into simple H
x
C-, H
x
N- or CN-
radicals and side chains [23]. The dissociated nitrogen (N) from pyrazine molecules is
In-situ formation of Aluminium Nitride composite powder


48
expected to react with highly reactive Al metal surfaces especially when Al is in
nanocrystalline structure after certain period of MM.

3.2 Experimental
Al powder (Alfa Aesar, -325 mesh, 99.5% purity) and pyrazine (H
4
C
4
N
2
) (Alfa Aesar,

99+% purity) were used as the starting materials in this study. 14g Al powder mixed
with pyrazine in the amount to satisfy the stoichiometric ratio of Al:N=1:1 was loaded
into 250 ml stainless steel vial together with 20 carbon steel balls in a 99.9% pure
argon atmosphere in an AMBRUAN glove box. A Retsch PM100 Planetary Ball Mill
was employed for MM at 300 rpm. The weight ratio of Al powder to ball was 1:20.
For blended powder mixture, Al and pyrazine were thoroughly mixed in an agate
mortar with an agate pestle.

After every 20h of milling, a small quantity of powder was withdrawn and annealed at
500 and 1250°C in a tubular furnace for one hour in purified argon flow. The blended
powder mixture was also annealed for the comparison study. An X-ray diffraction
(XRD) measurement using Cu K
α
radiation operating at 40 kV and 30 mA was carried
out for structural examination of MMed powders and MMed powders after annealing
at 500°C and 1250°C (designated hereafter as xxh-MMed, xxh-MMed-500 and xxh-
MMed-1250 samples respectively, where xx is milling hours and the blended powder
mixture is indicated as 0h). The microstructures of the samples were examined using a
Quanta 200F field emission scanning electron microscope (FESEM) and Jeol 2010F
TEM. Shimadzu DTG-60/60H was employed for the simultaneous measurements of
thermogravimetry and differential thermal analysis.

In-situ formation of Aluminium Nitride composite powder


49
3.3 Results and discussion
3.3.1 Structural evolution
Intensity (a.u.)
30 8040 7050 60

2

(degree)
Al
2
O
3
Al
500°C
1250°C

Figure 3.2 X-ray diffraction patterns of 0h-MMed Al-Pyrazine mixture after annealing
at 500°C and 1250°C for 1h.

Fig. 3.2 shows the X-ray diffraction patterns of 0h-MMed Al-pyrazine mixture
annealed at 500 and 1250°C for one hour in purified argon gas flow. Only Al peaks are
detected in the 0h-MMed-500 sample indicating no reaction between Al and pyrazine.
For the 0h-MMed-1250 sample, significant oxide formation is detected from XRD
patterns with high intensity alumina (Al
2
O
3
) peaks. Although annealing was carried
out in high purity argon gas, oxidation could not be fully prevented. Since the melting
temperature of pyrazine is 54°C which is far below the melting temperature of Al, it
evaporated at much lower temperature than that of reaction between Al and N. Hence,
AlN could not be formed.

The XRD patterns of the MMed aluminium-pyrazine mixtures at different milling
durations are shown in Fig. 3.3. All the diffraction peaks from the 20h- and 30h-MMed

samples are from the pure Al implying no formation of compound between N/carbon
In-situ formation of Aluminium Nitride composite powder


50
(C) atoms and Al. The DTA traces of MMed powder for different milling durations in
Fig. 3.4 show endothermic peaks at 668°C due to Al melting from the powders milled
for 20h and 30h. It is possible that the mechanical energy supplied to the milling
process was not high enough to break the bonds between C and N in the pyrazine
molecules and/or to overcome the diffusion barrier for N atoms to diffuse into the Al
particles.

Intensity (a.u.)
30 8040 7050 60
2

(degree)
AlN Al
2
O
3
Al
100h
20h
30h
40h
60h
80h

Figure 3.3 X-ray diffraction patterns of MMed Al-Pyrazine powder at different

milling durations.

As milling progressed, N atoms gradually diffused into highly defective Al particles.
After 40h of milling, AlN started to form as clearly shown in Fig. 3.3. Although weak
In-situ formation of Aluminium Nitride composite powder


51
Al peaks are detected, the amount of residual unreacted Al may not be large enough to
produce any endothermic peak corresponding to the Al melting in DTA traces (Fig.
3.4). It can be seen that the AlN peaks are broad, which corresponds to the
characteristics of nanocrystallites with the presence of interfacial components
composed of Al, H and N atoms [20]. The average crystalline sizes estimated using
Scherrer’s formula based on the theory of broadening of XRD diffraction peaks are 10,
9, 8 and 8 nm for 40, 60, 80 and 100h-MMed powders respectively.

DTA (mW/mg)
600
1400
2
00
8
00
4
0
0
Temperature (°C)
1200
1
00

0
0
2
-4
0
1
-2
-3
-1
Exothermic
100h
80h
60h
40h
30h
20h

Figure 3.4 DTA traces of the Al-Pyrazine powders MMed for different milling
durations.

Due to low melting point and asymmetric structure of pyrazine, it is unstable under
thermal treatment and/or mechanical activation [23]. Pyrazine therefore decomposed to
some extent after 40h of milling and a certain amount of released N reacted with Al
with the evidence of weak and broad AlN peaks. The proportion of AlN in the powder
increased with further milling. Higher and distinct AlN peaks indicating higher
In-situ formation of Aluminium Nitride composite powder


52
crystallinity of AlN particles could be observed in the 100h-MMed sample.

Disappearance of endothermic peaks at 668°C in DTA traces corresponding to Al
melting in the 60, 80 and 100h-MMed samples confirms the complete transformation
of Al to AlN and Al
2
O
3
.

Al
2
O
3
oxide peaks were detected in the MMed powders except in 20h- and 30h-MMed
powder samples. The sources of Al
2
O
3
formation in the AlN powder could have
originated from the oxide layers on the Al raw material [24], possible oxidation during
milling and powder handling due to its highly pyrophoric nature.

In the dry-milling process, under continuous impact and shear stress during MM,
Al
2
O
3
passivation layer on the as-received Al particles was fractured creating new
surfaces to react with free N dissociated from pyrazine ring structure. Due to the strong
affinity between Al and N, the N atoms were adsorbed on the newly created Al
surfaces and then incorporated in the interfaces by the pressure welding of the Al

powder followed by diffusing into the Al matrices through grain boundaries,
dislocations and other defects [25]. Milling induced dislocation and vacancy density in
the MMed Al enhanced the favorable situation for direct chemical nitridation
following the decomposition of pyrazine molecules. During mechanochemical
activation, increased diffusion of nitrogen was attributed to development of
nanostructural or amorphous phase [26]. The exothermic behavior owing to the
extensive heat of reaction enhances the reaction rate [27].

In-situ formation of Aluminium Nitride composite powder


53
30 8040 7050 60
2

(degree)
AlN Al
2
O
3
Al
100h
80h
60h
40h
30h
20h
Intensity (a.u.)

(a)

30 8040 7050 60
2

(degree)
100h
20h
40h
60h
80h
AlN
Al
2
O
3
Al
4
C
3
30h
FeAl
2
Intensity (a.u.)

(b)
Figure 3.5 X-ray diffraction patterns of MMed Al-Pyrazine powder at different
milling durations annealed at (a) 500°C and (b) 1250°C for 1h.

Figs. 3.5(a) and (b) show the XRD patterns of MMed-500 and MMed-1250 samples
respectively. The XRD patterns of MMed-500 samples are very much similar to those
of as-milled samples indicating the negligible effect of annealing at 500°C. No AlN

structure could be detected in 20h and 30h-MMed-500 samples. After annealing at
1250°C, AlN was found to form in both 20h and 30h-MMed powders. In addition to
AlN formation, oxide and carbide were detected from Al
2
O
3
and Al
4
C
3
diffraction
peaks. Al
4
C
3
is the only intermediate compound in Al-C binary phase diagram [28]. A
weak peak which might correspond to FeAl
2
was observed in all MMed-1250 samples.
In-situ formation of Aluminium Nitride composite powder


54
Al and Fe contamination from milling media formed FeAl
2
intermetallic during
milling. Compared to the powder annealed at 500°C, oxidation was more pronounced
and Fe contamination from milling media was clearly identified. Average crystalline
sizes of AlN after annealing at 1250°C were estimated to be 14, 13, 6, 9, 10 and 10nm
for 20, 30, 40, 60, 80 and 100h-MMed samples respectively indicating no significant

grain growth during annealing.

Table 3.1 Contents of C, H, N, Al and Fe under different conditions
Sample C (wt.%) H (wt. %) N (wt. %) Al (wt. %) Fe (wt. %)
100h-MMed-1250 28.83 0.50 17.86 43.98 1.08
80h-MMed-1250 28.16 <0.50 16.22 50.31 1.12
60h-MMed-1250 27.62 2.51 13.02 34.95 0.45
40h-MMed-1250 28.62 0.77 15.78 40.96 0.58
30h-MMed-1250 22.34 <0.50 11.26 45.35 0.86
20h-MMed-1250 17.58 0.51 8.92 52.87 0.84
100h-MMed-500 27.59 1.36 16.09 42.20 1.07
80h-MMed-500 26.88 1.36 14.18 41.08 0.79
60h-MMed-500 27.03 1.35 14.32 45.21 0.90
40h-MMed-500 26.41 2.33 13.49 43.73 0.73
30h-MMed-500 17.92 1.56 7.80 41.08 0.53
20h-MMed-500 15.97 1.37 7.34 50.55 0.65

From the results of chemical analysis as tabulated in Table 3.1, a certain amount of
carbon did exist in the MMed powders. However, only the 20h- and 30h-MMed-1250
powders showed the Al
4
C
3
carbide peak but not in 500°C annealed samples. After 20h
and 30h MM followed by subsequent annealing at 500°C for one hour, the mechanical
activation and thermal activation might not be sufficient for carbide formation. For
longer milling duration, the disappearance of carbide peaks might be due to the
emergence of carbide peaks with broadened AlN peaks due to grain refinement with
In-situ formation of Aluminium Nitride composite powder

55
longer milling duration and the carbide particles are too fine to produce diffraction
peaks.

3.3.2 Formation of alumina and AlN whiskers
Various microstructures of AlN have been reported in the literature such as fibers,
agglomerated particles, coarse granules, hexagonal whiskers, and faceted particles. In
the synthesized AlN products, different morphologies coexisted in most cases [29]. In
this study, the Al powders were subjected to micro-forging, fracture, agglomeration,
and deagglomeration during MM process and the crystals in the deformed powder
particles were heavily stressed and strained in a rather inhomogeneous manner [30]. It
is generally accepted that formation of protrusions or whiskers results from local
stress-relaxation process [31-33]. Since the stress and strain distribution in the MMed
Al powder is inhomogeneous and the fluctuations of growth conditions such as
localized distribution of absorbed nitrogen molecules, degree of Al vapor saturation
and reaction temperature, different morphologies of AlN whiskers and particles
coexisted in this present material.

After annealing Al-Pyrazine blend at 1250°C for one hour, white cotton-like Al
2
O
3

whiskers were observed to be surrounding the spherical shape Al melt in the middle.
TEM image with selected area electron diffraction (SAED) pattern (insert) in Fig.
3.6(a) reveals that the whiskers are alumina single crystal confirmed by the lattice
spacing of 3.5Å corresponding to diffraction from the (012) plane. Fig. 3.6(b), a high
resolution TEM (HRTEM) image of the lattice pattern from 0h-MMed-1250 sample,
shows an inter-planar spacing of 2.559Å corresponding to the (104) planes of the

In-situ formation of Aluminium Nitride composite powder


56
Al
2
O
3
crystal which appears to grow in a direction angled at ~30° with normal to the
(104) plane.


(a)

(b)
Figure 3.6 TEM images of (a) alumina whisker and SAED pattern (insert) and (b)
HRTEM image of lattice pattern from the same whisker (growth direction along black
arrow).

Fig. 3.7 shows the morphologies of 0h-MMed-1250 sample. In entangled alumina
whiskers, network-like structures with a radial outgrowth from a single stem-like
structure can be seen in Fig. 3.7(a). The alumina whisker has a (001) facet (C-facet)
that appeared as a flat section on the top with a simple six-sided structure as shown in
Figs. 3.7(b), (c), (d) and (e). Lack of droplet on the tip of the whisker excluded a VLS
mechanism and the possible growth mechanism of the alumina could be vapor-solid
(VS) route. From Figs. 3.7(d) and (e), a repetitive stacking of the platelets along the
growth direction gives rise to the complete structure. However, the existence of such
distinct layers disappeared while smooth appearance was significant with pillar-like
structure in Fig. 3.7(c). Such structure was resulted due to adatom hopping as a result

d
104
=2.559Å
In-situ formation of Aluminium Nitride composite powder


57

(a)

(b)

(c)

(d)

(e)

(f)
Figure 3.7. Morphologies of blended Al-pyrazine mixture annealed at 1250°C for 1h.


In-situ formation of Aluminium Nitride composite powder


58
of a reducing Ehrlich-Schwoebel (ES) barrier (step-edge barrier which accounts for the
diffusion barrier of an adatom over the edge of a step) and lateral growth became more
pronounced to generate near perfect uniform dimension [34]. When adatoms can

easily overcome the ES barrier and get down to the step-edge, layer-by-layer growth is
promoted. Multilayer growth mode starts to appear if there is an asymmetry in the
incorporation rate of adatom into the step and it is much lower than surface diffusion
[35]. Fig. 3.7(d) shows the incomplete crystal growth of alumina pillar with hexagonal
flat top oozing out from the surface.

Fig. 3.7(f) shows whiskers with kinks resulting from sudden change in their growth
orientation. It is unlikely due to whisker collapsing by their own weight since straight
whiskers up to 500 μm could be observed. It is possible that kinking may be caused by
the uneven flow of material across the growth interface [33].

Fig. 3.8 shows the FESEM images of MMed Al-Pyrazine mixture at different milling
durations ranging from 20h to 100h. Since Al is ductile material, the particles were
heavily agglomerated during MM and the aggregates were composed of particle size
ranging from micron to submicron. As reported by William [36], a broad particle size
distribution and the presence of large aggregates were observed. Refinement and
reduction in particle size with prolonged milling is evident in Fig. 3.8.

No whisker formation could be detected in all MMed samples annealed at 500°C. This
is because the annealing temperature is well below the melting temperature of Al to
have reaction between Al and N. The samples annealed at 1250°C consisted of both

In-situ formation of Aluminium Nitride composite powder


59

(a)

(b)

(c)

(d)

(e)

(f)
Figure 3.8 Morphologies of (a) 20h-, (b) 30h-, (c) 40h-, (d) 60h-, (e) 80h- and (f)
100h-MMed samples.
In-situ formation of Aluminium Nitride composite powder


60

(a)

(b)

(c)

(d)

(e)

(f)
Figure 3.9 Morphologies of (a) 20h-, (b) 30h-, (c) 40h-, (d) 60h-, (e) 80h- and (f) 100h-
MMed-1250 samples.

In-situ formation of Aluminium Nitride composite powder

61
AlN particles and whiskers as shown in Fig. 3.9. Figs. 3.10 (a) and (b) show a TEM
image with SAED pattern (insert) and HRTEM image of lattice pattern from 100h-
MMed-1250 sample. It reveals that the AlN particles are poly-crystalline with an inter-
planar spacing of 2.74 Å corresponding to the (100) planes of the AlN crystal grown
along the [100] direction.


(a)

(b)
Figure 3.10 (a) TEM image of 100h-MMed-1250 sample with SAED pattern (insert)
and (b) HRTEM image of lattice pattern from the same sample.

It is interesting to find that the formation of AlN whiskers was most abundant in the
30h-MMed-1250 sample followed by 20h-MMed-1250 sample. However, minimal
amount of whisker could still be seen in other milling durations. During milling, the
ductile Al particles were easily deformed plastically under the high impact kinetic
energy of the milling balls. This plastic deformation led to the formation of layered
structure in which pyrazine and/or nitrogen molecules dissociated from pyrazine and
the Fe particles from milling media (balls and vial) were entrapped.
Under direct
collision and relative friction during milling, the cold welded powders were fractured
from the milling media and the transfer of Fe atoms from milling tool to the powder
d
100
=2.74Å
In-situ formation of Aluminium Nitride composite powder

62
particles occurred [30]. These resultant Fe particles are very small in size probably in
the nanometer range. They formed FeAl
2
intermetallic with Al and played the role of a
catalyst in the vapor feed-liquid catalyst-solid crystalline whisker growth (VLS)
mechanism through which AlN whiskers grew with different morphologies.

The majority of whiskers could be described as long and straight filaments (length up
to 30 μm and diameter around 0.2 μm) with smooth surface in 20h- and 30h-MMed-
1250 samples. The droplet (a spherical ball) on the tip of AlN whisker (Fig. 3.9a)
indicates that the growth mechanism is VLS mechanism [37, 38]. High percentage of
Fe at the droplet from EDS (energy dispersive spectrum) analysis (Fig. 3.11)
confirmed the role of Fe as a catalyst.



Figure 3.11 The energy dispersive spectrum of droplet at the tip of AlN whisker.

From DTA traces (Fig. 3.4) and XRD patterns (Fig. 3.5), there was no indication of Al
existence implying complete transformation of Al to AlN and Al
2
O
3
after 60h-MM.
With prolonged milling duration, the grain size becomes smaller and a large volume
fraction of the atoms resides in the grain boundaries resulting in higher reactivity and
diffusivity [39]. Therefore, most of the Al is transformed into AlN, with only a few

whiskers being produced from the very small amount of residual Al in the samples
MMed for 40h or longer. Different from the whiskers in 20h- and 30h-MMed-1250
KeV
Counts
Al
In-situ formation of Aluminium Nitride composite powder


63
samples, without droplets at the tips, whiskers in bead-shaped and chain-like structure
with kinks and wavy shape are detected as shown in Figs. 3.9(c), (d), (e) and (f). AlN
whiskers with such structures have been reported by Jung et al. [40].

During milling, the dissociated N atoms from pyrazine were absorbed on the newly
created Al surfaces and then incorporated in the interfaces formed by the pressure
welding of the Al powders. Due to the inhomogeneous distribution of decomposed N
molecules in the Al layered structure and the distribution of stress field resulted from
milling, various morphologies of AlN whisker were detected in the 20h- and 30h-
MMed-1250 samples as shown in Fig. 3.12.

According to Wagner and Ellis [41], the VLS mechanism postulates that the presence
of a tiny liquid droplet can act as a preferred site for whisker growth from the vapor.
From Al-Fe binary phase diagram, at the processing temperature of 1250°C, the FeAl
2

would have melted and the AlN vapor dissolved into the FeAl
2
droplets at high
saturation pressure [40]. Process of precipitation of AlN begins leading to the growth
of AlN whiskers when the dissolved AlN reaches a certain supersaturation level.

Generally, a higher supersaturation would favor the formation of powders, whereas a
lower supersaturation ratio results in whiskers [42].

It is clearly seen in Fig. 3.12(a) that a secondary nucleation took place at the original
main AlN whisker. A number of AlN whiskers are branching out from the main
whisker with droplet at the tip. It could be due to the excess supersaturation of the
droplet, which activated secondary growth sites on the sides of the whiskers for lateral
growth, causing the very small ~0.2µm diameter side branches to occur [13,38].
In-situ formation of Aluminium Nitride composite powder


64

(a)

(b)

(c)

(d)

(e)

(f)
Figure 3.12. Various morphologies of AlN whiskers in (a,b,c) 30h- and (d,e,f) 20h-
MMed-1250 samples.

In-situ formation of Aluminium Nitride composite powder

65
Whiskers which had a structure as chains of roundish beads of c.a. 4µm in diameter
were detected as shown in Figs. 3.12(b) and (c). Similar whisker structure has been
reported by Jung et al. [40,43]. Liquid droplets serve to dissolve the dissolving vapor
and provide the sites for crystal growth by virtue of being in contact with the growing
surface [44]. The whiskers grew by diffusion of AlN into an FeAl
2
droplet. At the
same time, the whiskers also absorbed vapor phase AlN to grow in radial direction. At
times the vapor surrounding the whisker would be exhausted and saturation of AlN
vapor dropped momentarily. Subsequently, when saturation reached a sufficient level
through diffusion of AlN vapor, solid AlN precipitated from FeAl
2
droplet. Such
alteration of saturation level modulated the whisker diameter resulting in whisker to be
bead-shaped and chain-like.

Wavy or serrated AlN whiskers were observed as shown in Figs. 3.12(c) and (d). For
AlN whiskers originated at the interstitial pores between the welded Al layers, they
need to be extruded from the Al agglomerate and/or welded layers and break the AlN
surface layer because of the exclusively small diffusion coefficient of nitrogen in the
nitride layer. Due to the difference in thermal expansion coefficients between AlN and
Al metal, heat released from the exothermic reaction of Al and N affects cracking in
nitriding layer [45]. Those whiskers which underwent such protrusion process grew
without a specific direction resulting in wavy or serrated structure. On the other hand,
Cacerres proposed that the serrated morphology could be attributed to nonplanar
growth mechanism where two competitive densely packed planes, such as the basal
and pyramidal planes were active during whisker growth [46].

In-situ formation of Aluminium Nitride composite powder

66
Whiskers formed by stacking of AlN crystals of different sizes are clearly observed in
Figs. 3.12(e) and (f). As depicted in Fig. 3.12(f), the AlN crystal is in irregular cross-
sectional shape with steps at its side. Without a specific direction, AlN crystals are
joined face-to-face to form stack structure. Localized temperature rise might have
caused AlN to be more crystalline with faceted morphology and the variation of
growth conditions such as temperature fluctuation during furnace cooling, vapor
pressure in the isolated pores could also have caused chain-extrusion of AlN crystals
one by one in unspecific the growth direction.

3.4 Conclusions
An attempt was made to produce AlN composite powder via MM from the starting
materials of Al and pyrazine. It appeared that whiskers could not be grown in both as-
milled and blended samples at annealing temperature of 500°C which is well below the
melting temperature of Al. At 1250°C annealing temperature, alumina whiskers were
formed from Al-pyrazine blend as a result of oxidation during annealing. Most
abundant growth of AlN whiskers has been observed in the 20h- and 30h-MMed-1250
samples with highest content of unreacted Al. Therefore, it is suggested that the AlN
whiskers grew from the melt of residual Al in the milled samples.

MMed powders annealed at 1250°C are composed of both particles and whiskers of
AlN. Majority of whiskers in the 20h- and 30h-MMed-1250 samples are long and
straight filaments with droplet at the tip indicating the VLS growth mechanism. Fe
particles from contamination of milling media were believed to form FeAl
2
which
plays an important role as a catalyst. Various morphologies such as multidirectional
growth structure, bead-shaped and chain-like structure, serrated structure, and stack

In-situ formation of Aluminium Nitride composite powder


67
structure of AlN single crystals were also observed. Different microstructures were
developed with different growth mechanisms as a result of growth condition
fluctuations including the amount of residual unreacted Al, reaction temperature,
saturation level and internal strain distribution due to MM. The complex interplay of
growth phenomena in the process has made conclusive proposal of an overall reaction
mechanism difficult.

3.5 References
1. G Selvaduray, L Sheet, Mater Sci Technol 9 (1993) 463–473.
2. H Scholz, P Greil, J Mater Sci 26 (1991) 669–677.
3. SE Pratsinis, G Wang, S Panda, T Guiton, AW Weimer, J Mater Res 10 (1995)
512–520.
4. I Kimura, K Ichiya, M Ishii, N Hotta, T Kitamura, J Mater Sci Lett 8 (1989)
303–304.
5. C Sangural, Y Kinemuchi, T Suzuki, W Jiang, K Yatsui, Jpn J Appl Phys 40
(2001) 1070–1072.
6. AJ Chang, SW Rhee, S Baik, J Am Ceram Soc 78 (1995) 33–40.
7. AV Virkar, TB Jackson, RA Cuttler, J Am Ceram Soc 72 (1989) 2031-2042.
8. H Morkoç, S Strite, GB Gao, ME Lin, B Sverdlov, M Burns, J Appl Phys 76
(1994) 1363-1398.
9. B Forslund, J Zheng, J Mater Sci 28 (1993) 3132-3136.
10. JM Haussonne, J Losiec, JP Bertot, L Lostec, S Sadou, Am Ceram Soc Bull 72
(1993) 84-90.
11. R Fu, H Zhou, L Chen, Y Wu, Mater Sci Eng A266 (1999) 44-51.
12. S Yu, D Li, H Sun, H Li, H Yang, G Zou, J Cryst Growth 183 (1998) 284-288.
13. H Zhou, H Chen, Y Liu, Y Wu, J Mater Sci 35 (2000) 471-475.

14. AS Segal, SY Karpov, YN Makarov, EN Mokhov, AD Roenkov, MG Ramm,
YA Vodakov, J Cryst Growth 211 (2000) 68-72.
15. M Yano, M Okamoto, YK Yap, M Yoshimura, Y Mori, T Sasaki, Diamond
Relat Mater 9 (2000) 512-515.
16. PE Evans, TJ Dvies, Nature 197 (1963) 587-587.
17. H Iton, H Morikawa, K Sugiyama, J Crystal Growth 94 (1989) 387-391.
18. R-C Juang, C-C Chen, Mater Sci Eng A 458 (2007) 210-215.
19. R Fu, K Chen, X Xu, JMF Ferreira, Mater Lett 59 (2005) 2605-2609.
20. A Calka, JI Nikolov, Nanostruct Mater 6 (1995) 409-412.
21. JI Nikolov, JS Williams, DJ Llewellyn, A Calka, Mater. Res. Soc. Symp. Proc.
481 (1998) 649-654.
22. K Komeya, N Matsukaze, T Meguro, J Ceram Soc Jpn 101 (1993) 1319-1323.
23. WA Kaczmarek, BW Ninham, I Onyszkiewicz, J Mater Sci 30 (1995) 5514-
5521.
24. V Rosenband, A Gany, J Mater Process Technol 147 (2004) 197–203.
25. M Miki, T Yamasaki, Y Ogino, Mater Trans JIM 34 (1993) 952-959.
26. Y Ogino, S Murayama, T Yamasaki, J Less-Common Met 168 (1991) 221-235.
In-situ formation of Aluminium Nitride composite powder


68
27. O Kubaschewski, CB Alcock, Metallurgical Thermochemistry, Pergamon,
Oxford, 1979.
28. C Qiu, R Metselaar, J Alloys Compd 216 (1994) 55-60.
29. R-C Juang, C-J Lee, C-C Chen, Mater Sci Eng A 357
(2003) 219-227.
30. L Lü, MO Lai, Mechanical Alloying, Kluwer Acamdemic Publisher, Boston,
1998.
31. H Saka, K Tsujimoto, S Fujino, K Kuroda, H Takatsuji, S Tsuji. In: Stressed
Induced Phenomena in Metallization, ed. by H Okabayashi et al., pp. 371-381.

Fourth international workshop, Tokyo, Japan. 1997.
32. K Hinode, Y Homma, Y Sasaki, J Vac Sci Technol A 14 (1996) 2570-2576.
33. IA Blech, PM Petroff, KL Tai, V Kumar, J Cryst Growth 32 (1976) 161-169.
34. S-C Shi, S Chattopadhyay, C-F Chen, K-H Chen, L-C Chen, Chem Phys Lett
418 (2006) 152-157.
35. KH Hong, JK Yoon, PR Cha, Appl Surf Sci 253 (2006) 2776-2784.
36. W Rafaneillo. In: Carbide, Nitride and Boride Materials Synthesis and
Processing, ed. by AW Weimer, pp. 43-78. Chapman & Hall, London, New
York. 1997.
37. WG Miao, Y Wu, HP Zhou, J Mater Sci 77 (1997) 1969-1975.
38. JV Milewski, FD Gac, JJ Petrovic, SR Skaggs, J Mater Sci 20 (1985) 1160-1166.
39. C Suryanarayana, FH Froes, J Mater Res 5 (1996) 1880-1886.
40. W-S Jung, H U Joo, J Cryst Growth 285 (2005) 566-571.
41. RS Wagner, WC Ellis, Trans Met Soc AIME 233 (1965) 1053-1064.
42. M-J Wang, H Wada, J Mat Sci 25 (1990) 1690-1698.
43. W-S Jung, TJ Lee, B-K Min, Mater Lett.57 (2003) 4237-4242.
44. J Dong, W Shen, B Zhang, X Liu, F Kang, J Gu, D Li, N Chen, Carbon 39
(2001) 2325-2333.
45. T Okada, M Toriyama, S Kanzaki, J Eur Ceram Soc 20 (2000) 783-787.
46. PG Caceres, J Am Ceram Soc 77 (1994) 977-983.